Synthesis, characterization, and evaluation of antibacterial activity of transition metal oxyde nanoparticles

Nanoparticles (NPs) have a wide range of applications in various areas. For health application, cytotoxicity tests are used to ensure its efficiency and safety. In this paper, ZnFe2O4, CoFe2O4, Zn0.5Co0.5Fe2O4 NPs were synthesized, characterized and their antibacterial properties were evaluated. The Sol-Gel method was used to synthesize the NPs. Their electronic and crystallographic structures were characterized by Fourier Transform Infrared Spectroscopy Analysis (FTIR), X-ray fluorescence (XRF), X-Ray Diffraction (XRD), and Transmission Electron Microscopy (TEM). To perform the antibacterial evaluation, ferrites were dispersed through nanoemulsion to prevent the crystals from accumulating together. Then the evaluation was performed through microdilution in a 96-well plate and diffusion in agar disc in contact with 3 different strains of Staphylococcus aureus and Escherichia coli. It demonstrated that the Sol-Gel method was efficient to synthesize NPs with suitable sizes for health application. All synthesized NPs showed the inhibition of bacterias with different concentrations used.


Introduction
Nanotechnology is commonly known as a multidisciplinary science, involving the physical, chemical, and biological areas, where important applications are widely observed, such as in the industrial sector and in medicine [1]. The biological use of nanoscale materials, also known as nanoparticles (NPs), can be studied both in in vitro and in vivo [2,3]. Various applications of NPs are related to the optical properties, size, shape, surface aspects, etc [4].
Because of their small size, unique physical-chemical characteristics have been observed [5]. Some NPs present magnetic characters that are composed of transition metals oxides, presenting the formula of MFe 2 O 4 , where M is a divalent metal, such as cobalt (CoFe 2 O 4 ) and zinc (ZnFe 2 O 4 ) [6][7][8]. Ferrites can be synthesized by multiple methods, i.e., solid-state reaction [9], polymeric precursors [10], thermal decomposition [11], hydrothermal [12], etc. The polymeric precursor method involves a sol-gel process, which starts with precursors in the liquid state, followed by the formation of the sol phase, which is a colloidal suspension. For the gel phase, organic precursors that have a chelating function are usually used, which may be citrate, gelatin, etc. And then the final product is calcined to form stable and homogeneous crystalline structures [10].
The incessant search for new antimicrobial drugs is due to the large number of microorganisms resistant to conventional antimicrobials [13], where NPs are considered as one of the alternatives. The activities provided by metal ions in biological environments have stimulated the research and development of these compounds as therapeutic agents in the treatment of infectious diseases [14]. Ag NPs accumulated in bacterial membrane can cause perforations in the membrane and result in the death of cell [15]. Ferrite NPs have an important role against Gram-positive and Gramnegative bacteria, suggesting that they can be used as antibacterial agents [16]. Ferrites NPs have efficient antibacterial properties against different bacterias.
Some properties of NPs can influence nanotoxicity. Lima et al. [17] demonstrated that the smaller the size, the greater the effect of cell uptake and toxicity. The charge, composition, and morphology are important factors related to the interaction of these materials with cellular components. Other factors such as physical-chemical characteristics, crystalline structure, solubility, and agglomeration are related to the mechanisms of action that cause toxicity [18,19].

Characterization studies
In order to characterize the structures of ferrites, Fourier transform infrared (FTIR) spectroscopy and X-ray fluorescence (XRF) were used. To perform the FTIR, the sample was placed in a JASCO FTIR-4100 sample holder. Data acquisition was performed in the range of 400-4000 cm −1 with a step of 0.01 cm −1 . XRF was carried out on a Shimadzu model EDX-720. The measurements were performed in the range of 0-50 keV with a step of 0.01 keV. The experiments were performed at a pressure of 2.0 × 10 −8 mbar. The XPS spectra were collected using a conventional Al K α X-rays source with photon energy of 1486.6 eV. A VSW HA100 electron analyzer was used with 44 eV pass energy and 0.1 eV step. The base pressure in the analysis chamber was less than 5.0 × 10 −9 mbar. The binding energy (BE) scale was ajusted using the C 1s line at 284.6 eV as a reference. A 10°takeoff angle was used to increase the surface sensitivity of the core-level peaks. The data were analyzed using the Winspec software. Shirley backgrounds were subtracted from the experimental data results.
XRD was performed with a Rigaku X-ray Diffractometer. The samples were analyzed using the diffractometer with the following parameters: voltage of 40 kV, current of 20 mA, low incidence angle of 10°, θ-2θ mode scan, range of 20-70°, scanning rate of 0.5°/min, and Cu K α (λ = 1.541 Å) as source.
TEM was performed in a JEOL transmission electron microscope (JEM-2100) equipped with Thermoscientific EDS. The analyzed samples were prepared by adding 1 mg of the powder in an Eppendorf, where ethanol was used as a dispersing solvent and was placed in an ultrasound bath for 15 min. After dispersion, 250 µL of each sample was placed on the copper grid and rested on filter paper for ethanol evaporation. Then it was fitted in the microscope sample holder. The acceleration voltage used was 200.0 kV.

Dispersion of ferrites
Nanoemulsion composed by mineral oil, tween 80, propylene glycol, and water were used in the proportion of 1:1:2:6 to form a stable dispersion of ferrites in an aqueous medium. Each component was sterilized by autoclaving, filtration, and ultraviolet radiation.

Antibacterial tests
To perform the antibacterial tests, two methods were used: minimum inhibitory concentration (MIC) and disk diffusion on agar. Both tests were performed on three different strains of two species of bacteria: Gram-positive, Staphylococcus aureus ATCC (American Type Culture Collection) 25923, 29213, and 43300, as well as Gram-negative, Escherichia coli ATCC 25922, 51446, and 35218. Each strain was diluted in a saline solution, which was at the MacFarland scale 0.5,~1.5 × 10 8 UFC/mL [21].
The MIC method was performed to determine the lowest concentration capable of inhibiting the proliferation of the tested microorganisms. Thus, the NPs and strains were diluted in the following concentrations: 1 mg/mL, 0.5 mg/ mL and 0.25 mg/mL. The tests were performed using 96well plates with a "flat" bottom and in triplicate for each concentration of ferrite in MH broth. Three different types of inhibition controls were carried out. The positive inhibition controls were composed by Penicillin G Potassium at 10,000 UI/mL, and streptomycin sulfate at 10 mg/mL. The negative control was realized without the addition of microorganism. The microplates were incubated in an oven at 37°C for 72 h. The microplate spectrophotometer was used to read the optical density, using the 630 nm primary filter, in the periods of 24, 48, and 72 h [22].
The disk diffusion test was realized according to the Kirby-Bauer Method. The inoculum of the microorganism was seeded over the entire surface of MH agar in a Petri disc. Then the NPs of 2 mg/mL were deposited on the agar surface. The plates were incubated in an oven at 37°C for 24 h. The results were analyzed according to the presence or absence of a growth inhibition halo around the discs [20,[23][24][25]].

Determination of cell viability
The cytotoxicity of NPs was assessed by MTT assay. Briefly, Vero cells at a density of 1 × 10 5 cells/mL were seeded in a flat-bottomed 96-well microtiter plate (Kasvi-Brazil) and were incubated for 24 h at 37°C and 5% CO 2 , supplemented with 10% heat-inactivated FBS (Invitrogen, USA), 2 mM L-glutamine (Merck, Germany), and 100 U/ml penicillin and 100 μg/ml streptomycin sulfate (Sigma-Aldrich, USA). A range of concentrations from 62.5 to 1000 μg/ml of NPs was prepared using the cell culture medium DMEM and was added to the plate in triplicate. After 72 h, the treatments were removed and 100 μL of MTT reagent of concentration from 0.5 mg/mL in DMEM was added to each well and incubated for a further 2 h. The medium was then removed and 100 μL of DMSO solution was added to the wells. Finally, the plate was read at 550 nm by a microplate reader.  [5,27,28].

FTIR
The spectrum (f) shows the absorption peaks for Zn 0.5 Co 0.5 Fe 2 O 4 NPs calcined at 250°C (800°C). The peaks at 412 cm −1 (418 cm −1 ), 583 cm −1 (575 cm −1 ), 1090 cm −1 (1086 cm −1 ), 1403 cm −1 (1395 cm −1 ), 1635 cm −1 (1643 cm −1 ), and 3437 cm −1 (3465 cm −1 ) were associated with the chemical bonds between the atoms of Fe-O, Fe-O or Zn-O or Co-O, C=O, OH, OH, and OH, respectively. The first two peaks were attributed to the metal-oxygen bonds (Fe-O) in the tetrahedral sites, and the second peak was associated to the metal-oxygen bonds (Fe-O, Zn-O, and Co-O) in the octahedral sites. This is the typical characteristic of the reverse spinel and normal spinel for Co 2+ substituted zinc ferrite Zn 0.5 Co 0.5 Fe 2 O 4 , magnetic NPs [27,[29][30][31]. Figure 2 shows the XRF spectra for the ZnFe 2 O 4 and CoFe 2 O 4 NPs calcined at different temperatures. In the Fig. 2a, b, the components present at 6.2 keV and 7.1 keV correspond to K α and K β lines of Fe atom. The components present at 8.6 and 9.5 keV are due to K α and K β lines of Zn atom. In Fig. 2c, d, the components present at 6.9 keV corresponds to K α and K β line of Co atom. The average chemical composition obtained for ZnFe 2

XPS
Wide-energy range XPS (survey) spectra were collected for assessing the cleanliness state of the NPs calcined at 800°C are shown in Fig. 3. The atomic ratios determined by XPS are consistent with the CoFe 2 O 4 and ZnFe 2 O 4 NPs stoichiometry. There had been a significant amount of carbon at the surface, which was reduced by the calcination. The C 1s peak at 284.6 eV was used as a reference BE for calibration. Figure 3b shows the measured HRXPS (high-resolution XPS) spectra of zinc ferrite NPs. Deconvolution of the Zn 2p 3/2 peak was performed in the Zn 2p spectra as shown in Fig. 3b. The peak at 1021.4 eV was associated to Zn 2+  occupying tetrahedral sites and the peak at 1022.6 eV was associated to Zn 2+ occupying octahedral sites in zinc ferrite [32]. Figure 3c shows the Fe 2p core-level of ZnFe 2 O 4 NPs calcined at 800°C. The XPS spectrum has three components. The BE of 711.7 eV (Fe 2p 3/2 ) and 725.1 eV (Fe 2p 1/ 2 ) was associated to Fe 3+ ions in octahedral site, with corresponding satellite peaks at 719.8 eV (Fe 2p 3/2 ) and 733.8 eV (Fe 2p 1/2 ), respectively. The BE of 714.8 eV and 728.0 eV, was related to Fe 3+ ions in tetrahedral sites in Zn ferrite NPs [32]. Figure 3d shows the XPS for O 1s core-level of ZnFe 2 O 4 NPs calcined at 800°C. This spectrum is formed by three components presents at 528.3, 530.5, and 532.6 eV. The   [33][34][35]. Figure 3f shows the XPS spectra of Co 2p 3/2 core-level calcined at 800°C, which has three components with BE of 780.3, 783.3, and 787.3, and 783.3 eV were associated to Co 2+ ions in the octahedral and tetrahedral sites [36,37]. The peak at 787.3 eV was associated to the shake-up satellite peak of Co 2p 3/2 main line [36]. Figure 3g shows the Fe 2p core-level of CoFe 2 O 4 NPs calcined at 800°C. The XPS spectrum has 3 components. The BE of 711.4 eV (Fe 2p 3/2 ) and 724.6 eV (Fe 2p 1/2 ) refers to the Fe 3+ ions in octahedral site, with corresponding satellite peaks at 719.8 eV (Fe 2p 3/2 ) and 733.4 eV (Fe 2p 1/ 2 ), respectively. The BE of 714.8 eV and 728.0 eV, was associated to Fe 3+ ions in tetrahedral sites.
The asymmetrical complex peak of the O 1s XPS spectrum (Fig. 3h) can be decomposed into three components, with BE of 528.3, 530.5, and 532.6 eV. Similar to the ZnFe 2 O 4 NPs mention previously, the component present at 528.3 eV was related to oxygen of bulk structure and the components at 530.5 and 532.6 eV were associates to carbonate or hydroxyl groups chemically bound to surface cations of NPs [32,38]. These XPS results are in perfect agreement with the XRD measurements.  Table 1. Figure 4c shows the XRD diffractograms of Zn 0.5 Co 0.5 -Fe 2 O 4 NPs calcined at 250 and 800°C. The diffraction peaks at 36°, 43°, 57°, and 63°corresponds to (311), (400), (511), and (440) facet, respectively. For the sample calcined at 800°C, the diffraction peaks at 30°, 36°, 37°, 43°, 52°, 57°, and 63°were associated to (220), (311), (222), (400), (422), (511), and (440) facet, respectively. The change of the width in diffraction peaks also suggested that NPs calcined at high temperatures have a larger size, as shown in Table 1. Table 1 shows the crystallites sizes for the synthesized NPs. The crystallites sizes were determined by Scherrer equation. Table 1 shows that the crystallite sizes changed after the high-temperature calcination [35,38] and showed that higher calcination temperatures produce higher intense peaks, increased grain size, and better crystallinity. These authors obtained 19 and 86 nm of cristallite size for zinc ferrite non-calcined and calcined for a longer time. Lima et al. [17] showed that the size of the NP has a great influence on cytotoxicity, where Co NPs produced using the sol-gel method, have greater cytotoxic power when low calcination temperatures are performed. Figueiredo et al. [39], measured the cristallites size equal to 12, 10, and 10 nm for zinc, cobalt, and mixed ferrites synthesized using the combustion reaction technique, without undergoing by calcination treatment.  Fig. 6, which were elaborated from the frequency sizes from the measurement of NPs by using Image J software. To obtain the histogram a total of N = 50 different particles sizes were used. Subsequently, a particle size histogram was mounted using the Sturges method [40]. The width (W) was obtained from the relation: W = (Dmax − Dmin)/k, where k = 1 + 3.322 log(N). The histogram is fairly well modeled by a log-normal distribution, as shown in Fig. 6 for all the NPs.

TEM
ZnFe 2 O 4 sample calcined at 250°C (Fig. 5a) showed characteristics of amorphous material, with an average size of 10.4 nm (Fig. 5b). For CoFe 2 O 4 NPs calcined at 250°C shown in Fig. 5e were observed the formation of partial amorphous structures. The same behavior was observed in the XRD measurements. These TEM images indicate that the CoFe 2 O 4 and Zn 0.5 Co 0.5 Fe 2 O 4 NPs calcined at 250°C were relatively uniform, which were around 12.4 (Fig. 5f) and 9.2 nm (Fig. 5j) in diameter, respectively. The TEM images of ZnFe 2 O 4 , CoFe 2 O 4 , and Zn 0.5 Co 0.5 Fe 2 O 4 NPs that were calcined at 800°C are shown in Fig. 5c, g, k respectively. Here, the NPs sizes were larger compared to that calcined at 250°C, which were around 21.9 (Fig. 5d), 36.7 (Fig. 5h), and 31.4 nm (Fig. 5l)

Antibacterial tests
Inhibition tests using the MIC method showed, for all ferrites, that the highest inhibitory concentration is greater than 1 mg/mL. Thus, in Figs. 7 and 8, show the results with the concentrations of 1 mg/mL, 0.5 mg/mL, and 0.25 mg/mL according to the percentage of inhibition obtained using inhibition for the strains of S. aureus (n = 3) and E. coli (n = 3). This percentage was compared to the positive control for inhibition, which was a known antibiotic, and the negative control, which was the medium without the presence of inhibition. According to Fig. 7 Table 2 presents the values obtained for IC50 for the inhibition analyzes of S. aureus strains. Thus, MIC was considered to be greater than 1 mg/mL, with only the IC50  Table 3 presents the IC50 values for the analysis of inhibition of E. coli strains obtained by non-linear regression. At all incubation times, and both calcination temperatures, MIC values were greater than 1 mg/mL. The IC50 value ranged from 0.91 to 1 mg/mL for the three incubation times tested and the two calcination temperatures.
A study carried out with Fe 3 O 4 NPs in contact with the resistant species E. coli showed that the antibacterial effect promoted by ferrite in this bacterium is dependent on the concentration. In addition, the authors reported that the size of the ferrite was related to the antibacterial action [41]. A study carried out with zinc oxide, iron oxide, and Zn/Fe oxide, showed greater inhibitory action for NPs with a higher molar zinc concentration, obtaining greater inhibition for strains of S. aureus than for strains of E. coli [20,42,43] performed an antibacterial evaluation with Zn 0.5 Co 0.5 Fe 2 O 4 ferrite in MRSA (S. aureus Methicillin Resistant) and E. coli and were able to confirm greater inhibitory action for S. aureus. This inhibition decreased according to the incubation time, in 2 h there was an approximately 80% inhibition for S. aureus and 70% for E. coli and in 24 it decreased to about 65% and 50% inhibition.
The cytotoxicity of NPs on Vero cells was evaluated by MTT assay. Figure 9 shows the viability of Vero cells exposed to 125, 250, 500, and 1000 μg/mL of ZnFe 2

Agar diffusion test
Results on the antimicrobial activity of CoFe 2 O 4 , ZnFe 2 O 4 , and Zn 0.5 Co 0.5 Fe 2 O 4 NPs are rare in the literature. In this work, the antimicrobial activities of the synthesized NPs were tested against Gram-positive and Gram-negative bacteria. The antimicrobial effects of the NPs were qualitatively measured by performing agar diffusion test against all the test microorganisms. The results of zones of inhibition are shown in Table 4. The agar plate inhibition tests were performed by the disc diffusion method, as a way to confirm the inhibitory action of the ferrites prepared here. The disk inhibition tests, following the Kirby-Bauer method [44]. The absence of microbial growth around the NPs is an indirect measure of the ability of the NP to inhibit the growth. With a concentration of 2 mg/mL, was evident the formation of growth inhibition halos around the disks on the medium with the bacterium. After the dispersion using the nanoemulsion was possible to obtain clearer and evident inhibition halos for all synthesized ferrites, as shown in Fig. 10. Figure 10 shows a visual difference between the formed diameters by the inhibition halos created by zinc, cobalt ferrites, and Co 2+ substituted zinc ferrite calcined at 250°C and 800°C in contact with S. aureus strains. It was observed that the calcination temperature changed the diameter of the formed halo. Figure 10a, b shows the halos formed for ZnFe 2 O 4 NPs calcined at 250 and 800°C when in contact with S. aureus strains. The diameter measurement of the formed halo for ZnFe 2 O 4 NPs calcined at 250 and 800°C were 13.1 and 12.1 mm, as shown in Table 4. Figure 10c, d shows the halos formed for CoFe 2 O 4 NPs calcined at 250 and 800°C when in contact with S. aureus strains. The diameter measurement of the formed halo for CoFe 2 O 4 NPs calcined at 250 and 800°C was 9.4 mm and 8.7 mm, as shown in Table 4.
The last NP placed in contact with S. aureus strains in the Agar diffusion test was the Zn 0.5 Co 0.5 Fe 2 O 4 , as shown in Fig. 10e, f. In this case, the diameter of the halo formed did not change significantly. The diameter measurement of the formed halo for Zn 0.5 Co 0.5 Fe 2 O 4 NPs calcined at 250 and 800°C were 10.9 mm and 10.5 mm, as shown in Table 4. In general, the diameters of the halos formed for ZnFe 2 O 4 NPs were slightly larger for NPs calcined at low temperature. These measured results suggest that the crystalline structure of NPs calcined at low temperature was slightly more effective in inhibiting S. aureus. XRD data shows that at low temperature, calcined NPs have more characteristics of amorphous material.  The second step was to place the same NPs in the agar diffusion test in contact with E. coli strains. The procedure used was the same as for S. aureus strain. Figure 11a, b shows the halos formed for ZnFe 2 O 4 NPs calcined at 250 and 800°C when in contact with E. coli strains. The diameter measurement of the formed halo for ZnFe 2 O 4 NPs calcined at 250 and 800°C was 8.8 mm and 7.8 mm, as shown in Table 4.  Table 4.
The last NP placed in contact with E. coli strains in the Agar diffusion test was the Zn 0.5 Co 0.5 Fe 2 O 4 , as shown in Fig. 11e, f. The diameter measurement of the formed halo  Table 4. Similarly, the diameters of the halos formed for ZnFe 2 O 4 NPs were slightly larger for NPs calcined at low temperature. These measured results suggest that the crystalline structure of NPs calcined at low temperature was slightly more effective in inhibiting E. coli strains.
The validity of the agar diffusion method was verified using the classic antibiotic chlorophenicolc. It is used as a controls. Figure 12 shows the inhibition halos for bacteria S. aureus in contact with chloramphenicol antibiotic controls (A), negative control (distilled water in the disc) (B), and control of the dispersion used in the test (nanoemulsion deposited in the disc) (C). Was also compared the antibiotic chloramphenicol effect (D), negative control (E), and the nanoemulsion (F) in contact with the E. coli bacteria.
The results presented in the figures show that the nanoemulsion did not interfere significantly in the results of inhibition of the produced NPs, when tested with E. coli and S. aureus, as shown in Table 4. A study carried out with α-Fe 2 O 3 (hematite) also showed that S. aureus bacteria are more sensitive than E. coli in an agar diffusion test, which is due to the greater resistance of Gram-negative because it has an extra lipid membrane, which makes it difficult the passage of NPs [45]. The authors Mandal et al. [46], showed that the inhibition halos formed by zinc ferrites are larger for S. aureus than when testing with E. coli. In this study, an inhibition zone equal to 22 mm was obtained for S. aureus and 14.6 mm for E. coli. In general, our results for the sizes of inhibition halos are of the same order of magnitude as those obtained by Mandal et al. Another study whose results were similar to ours is with NPs of Mn 1 −x Ni x Fe 2 O 4 , the authors obtained, using the agar well diffusion method, halos equal to (18 mm) for S. aureus and smaller halos (12 mm) for E. coli [16]. Tests using the modified Kirby-Bauer diffusion method showed that for greater zinc molarities in the mixed cobalt and zinc ferrite, greater inhibition halos were formed [24]. Similar to this study, it was also observed in the present study that the halos formed for the cobalt ferrites were inferior to the halos formed for the other ferrites. Ashour et al. [47], in tests with Zn 0.5 Co 0.5 Fe 2 O 4 by disk diffusion, showed that strains of S. aureus compared to strains of E. coli showed a greater   E. coli, and other strains, showed that the tested ferrite showed inhibition potential for both classes of bacteria, as both are negatively charged, which favor interaction electrostatic between the cell and the NPs or ions released from them. Some factors, such as the composition and structure of the cell wall, may also be related to the rate of bacterial death. It is seen that the inhibitory capacity for S. aureus and E. coli bacteria are different. However, the structure of the bacteria can promote greater attraction to NPs, since E. coli consists of lipid A, lipopolysaccharide, and peptidoglycan; while the cell wall of S. aureus is composed mostly of peptidoglycan [23,[49][50][51]. Some studies suggest that the inhibitory action generated by NPs is due to the generation of reactive oxygen species (ROS), which end up causing bacterial death. Results by Arakha et al. [52] showed that the antimicrobial mechanism involved in bacterial death is through the release of ROS when in contact with iron oxide NPs, in which NPs with positive and negative surface charges were tested, obtaining greater antimicrobial activity to those positively charged. The mechanism generated by photocatalytic reactions causes oxidative stress in bacteria. These end up suffering a rupture of the cell membrane with the leakage of cytoplasmic materials [53]. Other possible mechanisms are the attraction of NPs to the bacterial surface, due to the differences in loads. Metal cations are attracted by the negative surface charge of bacteria, they accumulate allowing entry into the cell, causing pores in the wall/membrane and loss of intracellular content [54].

Conclusions
The synthesis of ZnFe 2 O 4 , CoFe 2 O 4 , and Zn 0.5 Co 0.5 Fe 2 O 4 ferrites was successful by the sol-gel method. FTIR measurements showed important peaks for the transition metal ferrites of Zn and Co. The analysis by XRF detected higher percentages of iron with proportions of 2:1 for both Fe:Zn and Fe:Co. X-ray diffraction made it possible to structurally characterize ferrites, confirming that the material is of the crystalline type and that the degree of crystallinity has increased for higher calcination temperatures (800°C). The micrographs obtained by transmission electron microscopy showed that the ferrites calcined at a lower temperature, 250°C were small in size compared to the ferrites calcined at a higher temperature, 800°C. The antibacterial activity of the three ferrites was partially shown for the tests at concentrations of 1 mg/mL, 0.5 mg/mL, and 0.25 mg/mL with MIC greater than 1000 µg/mL for the two bacteria tested. According to the halos obtained, for discs at a concentration of 2 mg/mL, it was concluded that there was greater inhibition efficiency in the order of ZnFe 2 O 4 > Zn 0.5 Co 0.5 Fe 2 O 4 > CoFe 2 O 4 . Thus, it is concluded that the synthesized ferrites have dose-inhibitory capacity dependence.